Daily Search Forum Recap: April 28, 2017

Here is a recap of what happened in the search forums today, through the eyes of the Search Engine Roundtable and other search forums on the web.

Search Engine Roundtable Stories:

Other Great Search Forum Threads:

Source: SEroundTable

6 key paid search trends from Merkle’s Q1 2017 report

Merkle’s Digital Marketing Report for Q1 points to strong growth from Google and the continued strength of PLAs (Google Product Listing Ads). Expanded text ads have yet to yield promises of CTR gold. Bing and Yahoo’s lack of mobile market share is hampering growth. Here’s a look at some of the key trends from the report. (Keep in mind the data reflects spending from Merkle’s own client base, which skews large retailer.)

AdWords Q1 year-over-year growth outpaced that of Q4

Spending on Google AdWords increased 21 percent year over year in Q1 2017, up from 19 percent in Q4 2016. Click volume increased 20 percent over the previous year. CPCs ticked up 1 percent.

Merkle credits the addition of a fourth mobile text ad, PLAs in image search, Google Maps ads and the return of separate device bids as key contributors to growth over the past year.

Tablet bids fall, mobile bids improve relative to desktop

Tablet bids have steadily declined relative to desktop since Google enabled advertisers to bid separately on the two devices last summer. Merkle says decoupling tablets from desktop helped drive growth, with advertisers able to adjust bids separately for higher-value desktop clicks.

Google AdWords phone and desktop spend increased 51 percent and 12 percent, respectively, while tablet spend fell 23 percent.

Phone CPCs continued to gain ground on desktop in Q1. For non-brand queries, phone CPCs were 43 percent lower than desktop CPCs in Q1, compared to being 51 percent lower in Q4. Tablet CPCs were 25 percent lower than desktop in Q1, down from near-parity in early 2016, when the devices were combined in bidding.

Source: Merkle

PLAs keep growing faster than text ads

With 52 percent click share, PLAs accounted for more than half of retail search ad clicks in Q1, up from 48 percent in Q4. For non-brand queries, PLAs drove a whopping 75 percent of all clicks for retailers.

Spending on Google Shopping rose by 32 percent year over year in Q1, compared to 12 percent for text ads. Growing impression volume on mobile is helping to increase PLA click share, and growth was largely driven by non-brand queries.

Source: Merkle

Search partners, which includes Google image search, accounted for 11 percent of PLA clicks for the quarter, similar to Q4’s share.

Local Inventory Ads gaining traction

In Q1, Local Inventory Ads (LIA) accounted for 19 percent of all Google Shopping clicks on phones. CTRs on LIAs are 19 percent higher than PLAs on phone and desktop. Not surprisingly, online conversion rates for LIAs, which are designed to direct store traffic, are lower than PLAs.

CTR boost for expanded text ads (ETAs) remains elusive

Merkle has consistently reported seeing mixed results since ETAs first came on the scene last year. The Q1 results for ETAs show a CTR lift compared to standard text ads only on desktop ads shown at the bottom of the page.

After accounting for device, keyword type, and ad location, there is still no clear evidence that Expanded Text Ads are producing consistently higher click-through rates than the legacy Google text ad format.

Source: Merkle; high-traffic ad groups with both formats active in Q1

Overall, text ad spending on Google increased by 12 percent. However, non-brand text ad spending rose 16 percent year over year. Merkle says the fourth mobile ad unit and the addition of ads in Google Maps has done more to buoy text ad growth than format changes.

Bing & Yahoo mobile troubles

Across Bing Ads and Yahoo Gemini, spend fell by 14 percent compared to the previous year, marking the fifth consecutive quarter of spend declines.

With Google as the default search option on Android and iOS devices, mobile weakness continues to be a considerable handicap for Bing and Yahoo. Google accounted for 97 percent of mobile phone traffic in Q1. Bing and Yahoo clicks made up 19 percent of desktop clicks for the quarter.

There is much more detail in the report, including on organic, social and Amazon. It’s available for download here.

The post 6 key paid search trends from Merkle’s Q1 2017 report appeared first on Search Engine Land.

Source: SEland

Search Buzz Video Recap: Google Owl & Updates, AMP Bugs & How Search Works

This week in search, Google documented their fake news efforts code named Project Owl. We covered a potential Google algorithm update, but it is unclear if there really was a big update. Google told publishers to noindex content from the AP, Reuters and other feed sources. Google released their new how search works site and with that shared some numbers of their search tests and launches. Google said they have no plans to boost the HTTPS ranking signal. Google said links from high traffic pages arenât any more valuable. Googleâs request indexing doesnât replace sitemap files. GoogleBot doesnât understand audio in podcasts. Google said they can check out SEO agencies that are scamming. Google seems to have AMP bugs in the search results, they also had an analytics issue with AMP. Google said patents and research papers are good but they donât mean they use it in their ranking algorithms. Google updated the how search works site. Google added more review filters to the local pack. Google is testing carousels in the hotel booking feature and they are also testing call to actions there. That was this past week in search at the Search Engine Roundtable.

Make sure to subscribe to our video feed or subscribe directly on iTunes to be notified of these updates and download the video in the background. Here is the YouTube version of the feed:

For the original iTunes version, click here.

Search Topics of Discussion:

Please do subscribe via iTunes or on your favorite RSS reader. Don’t forget to comment below with the right answer and good luck!

Source: SEroundTable

Google: 1,653 Search Launches With 9,800 Live Experiments & 130,336 Search Quality Tests

We all know Google does a lot of testing, in fact, I probably only cover less than 5% of the tests people email or send me that Google is running. I see tons and tons of UI tests, feature tests, color changes, ranking changes and more. So when Google sends things are constantly changing in search – they are not lying.

The issue is defining what are UI and feature changes versus core algorithm changes. I like to cover the ranking changes and then Google comes back and says we make thousands of changes to search every year. Of course they do, but what percentage of those changes are ranking specific? A much smaller number.

Anyway, on the new how search works site that Google just revamped, Google documented the raw number of those changes to the search results.

Glenn Gabe pulled the metrics from the page and posted them on Twitter. The page that has this is over here. In summary:

  • Google quality raters conducted 130,336 search quality tests this past year
  • Google ran 18,015 side-by-side experiments in search
  • Google tried 9,800 live traffic experiments which hit about 1% of the searchers
  • Google ultimately launched only 1,653 of them based on all those tests

Year, only 1,653 launched fully but again. those are both UI, features, search rankings and more.

Here is a GIF screen capturing these metrics so we have it for when/if Google changes these metrics for next time:


Forum discussion at Twitter.

Source: SEroundTable

Google Mobile Results Tests Dropping AMP Results Completely?

Google AMP

Earlier this week, we reported that Google was potentially testing dropping the AMP logo from AMP powered mobile pages in the mobile Google search results. Well, since then, I’ve been seeing random reports of Google not just dropping the logo, but also not linking to the AMP page, when it is both available, properly canonicalized and linking all correctly.

Watch this video of it in action from Sergey:

This is the same result, in a different browser, which kind of implies a bucket test on the other one, showing AMP working properly in the mobile results:

The SEO tool RankRanger also spotted this after I covered it originally and told me it was more than just the logo drop too. They posted on Twitter that “Google is showing less AMP results,” adding that even when it “appears at times to show non-AMP version of certain pages despite availability.” They then shared their tool showing a significant drop in AMP pages showing up:

click for full size

RankRanger pulled some code from their sample and showed how the AMP canonical is there but Google in this test is not recognizing it:

click for full size

What is interesting is that when I asked Google about it earlier this week, they kind of made me feel like I was off base here. They said “as you know we do tens of thousands of experiments in search every year but there are no such plans at this moment to change anything. The logo tells users that the page they will open will do so reliably and consistently fast.”

Gary Illyes also:

Well, something is up. Either it is a bug or a test but something is up.

Forum discussion at Twitter.

Source: SEroundTable

Google Quickly Fixed The AMP Analytics Bug

Yesterday, those publishers and webmasters who deployed AMP on their sites were freaking out that Google stopped sending traffic to AMP pages. They soon quickly figured out it was a Google Analytics bug in how it was capturing AMP pages.

The reports began coming in yesterday morning on Twitter, Reddit, WebmasterWorld and more and more places.

Google’s Paul Bakaus from the AMP team quickly responded to the issue saying they are working on a fix:

Then by 2:30pm or so yesterday, Google fixed the issue:

It seems the fix would only be a retroactive fix but it is fixed going forward.

Here is a chart from one who was impacted showing the AMP page traffic returning after the bug was fixed:

So all should be good now.

Forum discussion at Twitter, Reddit, WebmasterWorld.

Source: SEroundTable

Google Cheese Logo Shows How Marie Harel Makes A Camembert

Google Cheese Logo Marie Harel

Today on Google’s home page is a special Google logo, a Doodle, for the 256th birthday of Marie Harel. Marie Harel was the first to make a camembert, which was in 1791. The Doodle on Google’s home page takes you through the several steps on how to make the soft, creamy, surface-ripened cow’s milk cheese.

The interesting thing is that animal activists are supposedly upset about the Doodle. There is a thread in the Google Web Search Help forums with a furious message from one of them that reads:

Cows don’t joyfully carry a bucket up an idyllic hillside with a lovely young maid so they can be milked and their milk made into cheese. If google’s content creators aren’t aware of the horrific standard living conditions that factory farming inflicts upon dairy cows, perhaps a little bit of research is in order before making absurdly ignorant cartoons about it?

Google wrote on their Doodle description:

Our Doodle celebrates Harelâs 256th birthday with a slideshow that illustrates how camembert is made, step by step. It’s drawn in a charming, nostalgic style reminiscent of early 20th-century French poster artists, such as Hervé Morvan and Raymond Savignac.

You can play with the Doodle slides over here.

Forum discussion at Google Web Search Help.

Source: SEroundTable

Daily Search Forum Recap: April 27, 2017

Here is a recap of what happened in the search forums today, through the eyes of the Search Engine Roundtable and other search forums on the web.

Search Engine Roundtable Stories:

Other Great Search Forum Threads:

Source: SEroundTable

What your paid search KPI says about you

There are plenty of options for particular metrics to drive paid search programs toward, but not every KPI is created equal. Here, we’ll dive into what your preferred KPI says about you and provide recommendations to help you assess whether your KPI reflects the goals you care about.

All insights are at least slightly hyperbolic and assume you have a one-track mind for your KPI of choice. I bet some of you out there can keep two, three, maybe even more KPIs in mind when managing accounts, as you should. You’ll just have to meld the descriptions to uncover those deep insights into your psyche that you’re reading this to find.

Click-through rate

You waaaaaaant them to want you.

You neeeeeeed them to need you.

Some brands care a lot about how likely a user is to click on an ad, and for good reason. Click-through rate (CTR) is a key component of Quality Score, which Google uses to determine both how well an ad ranks and what price an advertiser must pay for clicks. Higher CTR means better rank and lower cost per click (CPC).

Thus, brands should make their ads as appealing as possible to searchers. However, chasing CTR above all else can lead to some poor choices.

Could I probably make a few friends at a college activity fair with this sign? Sure.

Will they still be my friends if I don’t actually have the goods? Probably not.

Similarly, advertisers might be able to draw more clicks with some versions of ad copy but end up hurting conversion rate or brand sentiment in the process. As such, it’s important to avoid pitfalls like overpromising or misleading searchers in the pursuit of higher CTR.

Impression share

You love to be seen, and ensuring your ads are present for a minimum share of relevant searches makes you feel good about your competitive position in search. Consistently getting in front of users searching for queries related to your business offerings is important, and that visibility can sometimes come at the cost of less flashy metrics like return on ad spend.

Advertisers can access estimates for the share of relevant impressions their ads are showing on the SERP by using Google Auction Insights reports. These reports also feature information on how often specific competitors are showing ads, as well as other metrics such as average position and top-of-page rate.

Shooting for impression share in search as your primary KPI isn’t that dissimilar from a television ad buy focused on reaching a specific segment of the television-watching population. Maybe in search you’re focused on getting in front of 100 percent of people searching for “red running shoes,” while on television you want to reach ESPN viewers in the 7:00–8:00 p.m. time block on the East Coast, with different categories of keywords somewhat analogous to different television channels in terms of audience differences.

The difference is that the effects of online interactions with ads are, for the most part, much easier to track than the effects of television ads. Users can be cookied, and online orders and other interactions such as email signups properly attributed to ad clicks. Thus, shooting exclusively for impression share requires abandoning this information in pursuit of visibility.

For many brands, performance metrics such as return on ad spend (ROAS) are deemed more meaningful in allocating search spend as effectively as possible. However, there are some advertisers who would prefer to lock up a percentage of impressions related to particular categories of search, regardless of how ROAS turns out.

Return on ad spend

You care about profitability — unimpressed by all the impressions and clicks your ads are getting, you care about how much they cost and what your business gets out of them. Paid search is great for you, because it’s easy-peasy for most advertisers to track events such as orders and information inquiries to the clicks driving them.

The trickiness comes in how you measure that ROAS. Are you using last-click attribution? Accounting for offline interactions such as phone calls or store visits? Baking in the interactions that occur on a different device from the one searched on originally (cross-device)?

Not all ROAS measures are created equal, and the better an advertiser can get at attributing value to ads, the more effective optimization efforts will be. But since you care about performance so much, I bet you’ll put in the work to get attribution just right.

New customer acquisition

You’re all about growing your customer base, reaching those shoppers you’ve never been able to pull through other marketing channels. Maybe your business is such that paid search clicks from existing customers don’t carry much incremental value, or maybe you just really want to expand your market share.

There are a number of ways you can go about targeting those new customers. One is to shoot for cost per new customer acquisition as the primary KPI of search campaigns, while another strategy might involve trying to turn ads off to all existing customers via audience exclusions using Remarketing Lists for Search Ads (RLSA) and Customer Match audiences.

Choosing between these different methods of focusing on new customers should include evaluating what bringing back existing customers is worth. Is there literally no value to these interactions (very rare), or is it just that interactions with new customers should be valued more than those with existing customers (more common)?

Everyone likes new customers, but if you’re going to use new-to-file acquisition as a key measure of success, it’s best to make sure you’re not neglecting valuable existing clientele in search out of a desire for those shiny new shoppers.

Average position

I suppose what kind of person you are depends on what average position you want to be in. Want to stay in first? You’re a winner at heart who wants be number one. Really just hoping for position three? That’s a losing attitude right there.

Really though, shooting for average position in general is kind of a losing attitude. Sure, it might make sense to try to stay in position 1 for your own brand keywords, but shooting for a specific spot on the page with non-brand keywords regardless of cost and return on ad spend is pretty risky given the competitive nature of AdWords auctions. Get in a bidding war with a couple of competitors, and the ad spend can add up pretty quickly.

While there are some studies and individuals out there proclaiming that specific positions on the page magically perform better than others in metrics like conversion rate, most of the data and studies behind those conclusions can be politely characterized as pseudoscience. What’s more, even if those conclusions were true, they would change with every SERP update, such as the removal of text ads on the right rail and the addition of a fourth text ad above organic results rolled out in early 2016. As such, we recommend taking such declarations with a grain of salt.

Look at it all, focus on what’s important

The awesome part of paid search is that you can measure ALL of these metrics at the same time. The slight downside is that you have to choose what’s most important to your brand, which can be harder for some to pin down than others.

Paid search ads almost certainly have an impact on brand awareness and recall, though those effects can be hard to measure. Thus, it makes sense to take into account metrics like impression share to see how visible ads are for particular keywords.

At the same time, impressions don’t directly put money in the bank, and focusing on metrics like return on ad spend and cost per new customer acquisition can better ensure brands are getting enough out of their spend to justify the expense.

In the big scheme of things, brands should keep track of all of these metrics (as well as others, such as conversion rate, average order value and more) but optimize toward those that prove the real value of search.

The post What your paid search KPI says about you appeared first on Search Engine Land.

Source: SEland